The aim of this research was to examine individual physicochemical properties of grains of different malting and feed barley varieties using standard methods, as well as to determine morphological properties of grains using computer image analysis and to determine their possible use in predicting physicochemical properties of barley grains. In the research, 27 samples of barley were used, of which 19 samples were two-row, while 8 samples were multi-row varieties. Seven samples were classified as spring and and 20 samples as winter barley. Standard methods were used to determine the protein content, 1000 kernel weight and test weight, grain size fractions above 2.5 mm and grain vitreousness, and digital image analysis was used to determine the geometric characteristics of grain and grain cross-section. Area, perimeter, maximum and minimum Feret diameter, roundness, and the mean grey value were determined using the ImageJ program. Whole grain roundness was a significant predictor of protein content, and grain cross-sectional area was a predictor of the 1000 kernel weight. The area, length and roundness of the whole grain could be successfully used to predict the test weight of barley, and the predictors of grain size fraction above 2.5 mm were the length and roundness of the whole grain and the cross-sectional area of the grain. Vitreousness could be successfully predicted by determining the mean grey value of the grain cross section. Based on the correlation and regression analysis, it can be concluded that by applying digital image analysis in determining the geometric characteristics of the grain and its cross section, individual physicochemical properties of barley can be successfully predicted.